Utility of Reservoir Characteristics to Determine Minimum Sampling Effort Needed to Assess Sport Fish Populations in Kansas Reservoirs

Abstract

Effective reservoir fisheries management requires fish samples suitable for addressing objectives. These samples are typically attained with fish sampling gears using standardized protocols. Some standardized protocols for sampling reservoir fishes promote objective-based sampling, and many include a minimum number of gear deployments. This minimum number is often a function of reservoir surface area. However, reservoir size may not adequately predict the number of gear deployments needed to reach sample objectives. We used multiple linear regression to determine the relationships between the biological, chemical, and physical characteristics of reservoirs and the number of gear deployments needed to reach two sample objectives: (1) collecting 100 stock-length fish (N100), and (2) attaining relative SE of stock-length catch per effort (CPE) ≤25% (RSE25). These analyses were conducted using data from 34 Kansas reservoirs and six fish species. We used Akaike model averaging from a confidence model set (ΔAIC ≤ 2) to develop an average model for each sample objective (N100 and RSE25) and target species. Recent CPE was identified in N100 models for six target species and in RSE25 models for five. Conversely, reservoir surface area was identified in only one model for N100 and two for RSE25. These results suggest that reservoir characteristics other than surface area, particularly recent CPE, should be considered when developing minimum sample sizes for objective-based fish sampling protocols. This approach would allow managers to most efficiently allocate sampling effort while still collecting representative and precise fish samples.

Received March 6, 2015; accepted November 30, 2015 Published online March 22, 2016

Acknowledgments

We thank K. Austin, C. Bever, J. Goeckler, S. Lynott, D. Nygren, R. Marteney, and S. Steffen for thoughtful discussions associated with this project. Helpful comments from L. Kowalewski, K. Pope, and J. W. Schlechte improved this manuscript. We also thank all KDWPT personnel that collected field data from 2010 to 2013. Funding for this study was provided by KDWPT. Fish sampling was supported by Federal Aid grant F-22-R.